Getting precision timing right for financial services and the cloud
Author: Simon Kenny, CEO, Hoptroff
Daily life has changed a lot in the past year. Decades of innovation have occurred in mere months, as financial services worldwide have responded to the recent pandemic.
Now, it is widely accepted that the future will be one of hybrid working. Like everywhere else, financial services are hastily adopting technologies to support this change. At least 68% of global CIOs now state that “migrating to the public cloud and/or expanding private cloud” is a top driver of IT spending, in contrast to 48% in 2019.
In the past, moving stock exchange services to the cloud has been complicated by a lack of cloud-based regulatory technology that is as robust as the existing on-premise solutions. But as technology catches up, advanced network-delivered solutions, such as those for precision timing, are set to increasingly replace outdated hardware mechanisms.
Traceable time and regulation in the financial services
Ensuring highly accurate and traceable time is a key area that will need to be considered as financial services complete their transition to the cloud. Under MiFID II and CAT financial regulations, time synchronisation remains tightly controlled. For example, MiFID II requires every server that is an active market participant to have a maximum divergence of between 100 microseconds or 1 millisecond (depending on the speed of the server’s trading activity) from the benchmark of UTC (Universal Time).
The regulations address the widespread problem that clocks drift on any electronic device unless they are periodically corrected.
Although most people don’t think about it, time synchronisation is critical to the efficacy of financial services. Organisations’ technological estates are becoming more widely distributed, with more complex automated decision-making processes, and higher latencies between machines. Because of this, there is a greater need to ensure all devices in a distributed ledger process share the same time.
If they don’t, the timestamps in the record will put events in the wrong order and the implied latencies between events will be incorrect. This limits the utility of machine records and compromises data quality. It would not be possible to use the machine records reliably for the reconstruction of events. Equally, the records could not be used to report to auditors or regulators, nor to hold suppliers of services to account.
Time synchronisation in the cloud
Until recently, it had been extremely difficult to deliver fully compliant synchronised timing to the cloud. Instead, firms had to manage it all in house. This meant relying on expensive hardware architecture. In practise, each machine location required the installation of a GNSS antenna, cables, clocks, in addition to synchronization and monitoring software.
With the shift to the cloud, firms will need to adopt a resilient cloud network architecture. If the connectivity to a Primary Time Source (e.g. GPS, Galileo, or a standards institute) and grandmaster clocks can be moved to a secure cloud location, this can become a reference “Cloud Timing Hub”. This means it will simultaneously compare multiple timing sources to confirm the time is always correct. In effect, the solution will maintain timing accuracy and traceability to a higher standard than local installations that generally rely on one connection to primary UTC.
By adopting a cloud network architecture, users would connect to the “Cloud Timing Hub” via their choice of networks, using a software time synchronisation client that would steer the local device clock and keep a record of the adjustments. This removes the need to maintain local hardware installations wherever you have servers.
Here, the cost of traceable time implementation and maintenance is significantly reduced. The software will manage the periodic connections to the local machine clocks that will maintain time compliance thresholds and keep logs for future reference.
New traceable timing solutions
As migration to the cloud continues to boom, tech companies have responded by developing exactly this solution. Traceable Time as a Service (TTaaS) is a leading timing service that can deliver time to both on-prem location and the cloud.
The solution is fully integrated with major existing connectivity providers so that it can deliver traceable timing to any trading venue or major co-locations. The TTaaS software client can connect back to Timing Hub to establish a chain of comparisons to UTC and then steer the clock on the virtual server, so that any application timestamps from activity in the cloud will be fully verified, as per the leading financial regulations.
Notably, TTaaS can be tailored to suit different needs. With a ‘tiered’ product offering, users can A) choose an end-to-end complete solution that combines a traceable time-feed with synchronization, monitoring and reporting tools, and B) effectively only pay for the specific services they need. For example, a stock exchange may initially only want to buy a time-feed and continue to use their existing time synchronization and MiFID II regulatory reporting solution. This is ideal for users that already have one or several parts of a timing solution.
Resilient timing and GPS Satellites
In addition to their cloud capabilities, these new fintech solutions can successfully combat problems that users of hardware timing have historically faced.
Satellite failure continues to be a very real threat that will have enormous consequences on global critical infrastructure. If a satellite’s signal is spoofed or jammed, the time could become unreliable and in breach of financial regulations. This is such a large problem that last year, the United States government released an Executive Order to ‘Strengthen National Resilience through Responsible Use of Positioning, Navigation, and Timing Services’.
Because TTaaS is delivered via secure networks, it cannot be jammed or disrupted. This resiliency guarantees uptime in high performance environments such as financial services. Firstly, it can be used to replace old GNSS-based architecture completely, once the existing grandmaster clock has come to the end of its useful life – which occurs about every five years. In the meantime, it can provide disaster recovery capability for systems that are entirely dependent on GNSS.
Firms across financial services have long understood the importance of connectivity latency in maintaining a competitive trading infrastructure. Hardware timestamping infrastructure has been built up to measure response times between trading systems and venues. Now that business is migrating to a cloud-based infrastructure, there is a need to extend the verifiable time much more widely across the enterprise to all machines in a process, so that a verifiable ‘Time Fabric’ is established that will identify latency variations by machine and provide auditable timing logs. New solutions like TTaaS offer an architecture that can extend traceable time anywhere to support modern cloud architecture while maintaining compliance and resilience.
Ready to learn more?
When thousands of transactions take place every second this level of accuracy and reliability is required to give financial services, banks, hedge funds and fintech companies confidence that their transactions are being properly handled.
A highly accurate timing solution like the one outlined above is ready to be rolled out without the purchase and installation of additional timing infrastructure. Hoptroff Traceable Time as a Service (TTaaS®) is a range of network and software-based timing solutions that are simple, resilient, and cost-effective.
Whether you need the security of verifiable time for compliance, or sub-microsecond delivery into your data centre, our obsession with accuracy will transform your business.
This article was originally published in Finance Derivative on 8 April 2021.
